Mid-upper-arm circumference development and its validity in assessment of undernutrition

Transcription

1 Asia Pacific J Clin Nutr (1998) 7(1): Category OA 9 EN Mid-upper-arm circumference development and its validity in assessment of undernutrition Le Thi Hop 1 MD, R Gross 2 PhD, S Sastroamidjojo 3 MD, Tu Giay 1 MD, and W Schultink 2 PhD 1 National Institute of Nutrition, Ministry of Health, Hanoi, Vietnam 2Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH, Germany 3SEAMEO-TROPMED Regional Centre for Community Nutrition, University of Indonesia, Jakarta, Indonesia As part of a prospective, longitudinal study that was carried out in Hanoi, Vietnam (as from 1981), mid-upperarm circumference (MUAC) measurements, and the weight and height of children were taken from children for the first year of life. These measurements were recorded monthly, months: 3-monthly, months: 6- monthly and yearly for children beyond 72 months. The aim of this study was to observe development of MUAC of children on a longitudinal basis. Mid-upper-arm circumference increases by about 1 cm for boys and 1.5 cm for girls between 1 and 5 years. To recognise malnutrition, the cut-off level of MUAC 13.5 cm gave high values for sensitivity and specificity only for children aged 6 12 months. The cut-off levels of 14.0, 14.5 and 15.0 cm were more appropriate for children aged 13 24, and months. These results suggest that a single cut-off point of MUAC 13.5 cm cannot be used in screening moderate malnutrition for all children under age 5 but should be elevated with the increasing age of children. Key words: cut-off levels of mid-upper-arm circumference, mid-upper-arm circumference, nutritional assessment, sensitivity, specificity. Introduction The arm contains subcutaneous fat and muscle mass. In developing countries, where people may undergo a reduced food intake, lower subcutaneous fat and muscle mass tend to parallel decreased mid-upper-arm circumference (MUAC) and as a result are useful in the diagnosis of undernutrition. 1,2 It is an important tool in screening for undernutrition especially when weight and stature measurements are too demanding or are impossible, such as in emergency situations, and when the precise age of the child is unknown, since it has been suggested that MUAC is relatively independent of age for children between 1 and 5 years. 3 Arm circumference indicators of nutritional status have advantages for field use because of their simplicity and low cost; 4 6 while weight-for-age and weight-for-height indicators require an accurate portable scale and knowledge of age, as well as a reference table for field use. The MUAC cut-off points that are used in order to distinguish between normal and malnourished children vary. In some cases, a single cut-off point was chosen for children aged 1 5 years. 7,8 In other studies, a series of cut-off points were used to classify degrees of malnutrition. 9 The validity of MUAC indices in the assessment of nutritional status is still disputed. Some investigators claim that MUAC can differentiate normal children from those with malnutrition, 9 11 whereas others contest this finding. 12 Most studies of MUAC indicators were cross-sectional. The objective of the present study is to investigate the development of MUAC on a longitudinal basis. This paper also uses the concepts of sensitivity and specificity 13 in evaluating the performance of MUAC measurements as a screening test for identifying children judged as malnourished by the weight-for-age (W/A) and weight-for-height (W/H) indices. Subjects and methods Subjects A prospective, longitudinal study has been carried out in Hanoi from The study design and data collection was implemented by staff members of the National Institute of Nutrition in Hanoi. Two of the four districts of Hanoi were randomly selected from which maternity hospitals were identified. Newborn infants from selected maternities fulfilling the following criteria: (i) birth weight not less than 2500 g, (ii) normal and single birth, (iii) without physical abnormalities, (iv) Viet ethnic group, and (v) parents with normal health status were chosen to participate. There were four cohorts (year of birth): 1981, 1982, 1983 and In 1981 and 1982, 90 newborn infants were selected for each year and in 1983 and newborns were selected. A total of 31.7% of children (n = 96) up to the age 60 months did not complete the study due to changes in housing location. Some 205 children (123 boys and 82 girls), who were regularly followed up to 5 years of age, are included in this report. Data collection The children were followed up at their home. Anthropometric measurements of the children were conducted monthly Correspondence address: Dr Rainer Gross, GTZ/SEAMEO- TROPMED, PO Box 3852, Jakarta 10038, Indonesia. Fax: gtzseame@indo.net

2 66 LT Hop, R Gross, S Sastroamidjojo, T GiaY, W Schultink from birth to 12 months. From 12 to 36 months the children were monitored every 3 months and from 36 to 72 months measurements were taken every 6 months. After that age data collection was carried out on a yearly basis. The methods of measurements were followed according to the recommendation of WHO. 14 The date of monitoring was the birth date ± 4 days for the first 3 years and the birth date ± 10 days for the children aged more than 3 years old. A plastic tape was used to measure the MUAC. The midpoint of the left upper arm was determined by measuring the length from the tip of the shoulder to the tip of the elbow and the length was divided by two. The tape was wrapped around the straightened arm at that midpoint and the tape inspected to ensure it was neither too tight nor too loose. The measurement was recorded to the nearest 0.1 cm. The weights of the children were obtained using a Testut weighing scale (Testut, Paris, France) with a precision of 0.01 kg. The scale was controlled twice per day with a known weight (5 kg) and the subject was weighed with the minimum of clothing. The weights were recorded to the nearest 0.1 kg. The recumbent lengths of the children less than 24 months were measured using the babyboard (UNICEF, Copenhagen, Denmark) and the measurements were read to the nearest 0.1 cm. A microtoise (UNICEF) was used to measure the height of children (24 months) and the measurements were also recorded to the nearest 0.1 cm. All data were recorded for each child. Statistical analysis Data analysis was carried out by using the computer program EPI-INFO (version 6.01) and statistical Package for Social Science (SPSSPC+; SPSS, Chicago, IL, USA). Anthropometric indicators W/A and W/H were compared to the National Centre for Health Statistics (NCHS) reference data. 15 The cut-off point of 2 Z-scores (SD) was used to classify the nutritional status of the children. The MUAC increment 16 was used to assess the development of MUAC of the children followed up. The analysis of variance (ANOVA) was used for comparison between groups. Student s t-test was used for comparison of the means of two groups. The difference in prevalence data between groups was tested using chi-squared or Fisher s exact test. Pearson s correlation coefficient was also used for testing correlation between variables. Sensitivity and specificity results were calculated using Z-score < 2 W/A and W/H to define children truly malnourished (Table 1). Sensitivity was defined as the percentage of true positive for malnutrition that were identified when the MUAC was below the given threshold. Specificity was defined as the percentage of true negative for malnutrition that were defined when the MUAC was above the given threshold. Analysis of collected data was carried out in the SEAMEO-TROPMED Regional Centre for Community Nutrition, Jakarta, Indonesia. Ethical considerations Collected data have been used for study purposes only. The parents of the children followed up were informed about the purpose of the study and of the name of research institution before agreeing to participate. Assurance was given that their cooperation was voluntary and that no negative consequences would result to those who decided not to participate in the study. Also, the parents were informed that they could skip any question they did not want to answer. Study protocol was approved by the Ethical Committee of the National Institute of Nutrition in Hanoi. Results Table 1 details the MUAC data measured in boys and girls up to 10 years of age. It was found that the MUAC of the boys from 6 to 36 months of age increased only very slightly (0.4 cm). At the age of 60 months MUAC was 1 cm higher compared with that of the boys at 36 months. Similar but less marked results were found among the girls. The MUAC of girls was smaller than that of boys for all ages from birth to 9-years-old. After this age MUAC was almost the same for both sexes. Mid-upper-arm circumference gain during the first year was very high (3.49 cm for boys and 2.99 cm for girls), then followed by very low increments of MUAC (< 0.5 cm/year from 2 to 10-years-old). At the age of 12 and 13 years increased growth rate in MUAC was observed and its standard deviation became larger for male and female children (Fig. 1). Table 1. Development of mid-upper-arm circumference (cm) by age and sex of the children compared with National Center for Health Statistics (NCHS) reference standards quoted by Gibson (1990) Boys Girls Age Mean ± SD Mean ± SD (months) n (% of NCHS) n (% of NCHS) ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

3 Mid-upper-arm circumference in assessment of malnutrition 67 Table 2 indicates the prevalence of moderate malnutrition for the cut-off point (MUAC < 13.5 cm) of children from 6 to 60 months on a longitudinal basis in relation to the W/A and W/H indices. Mid-upper-arm circumference gave a higher prevalence of malnutrition in the first year (6 12 months) when compared with W/A and W/H; it then decreased with increasing age. From 12 to 60 months the prevalence of malnutrition indentified by W/A were higher than that of MUAC and W/A. Sensitivity and specificity of the selected cut-off levels of the MUAC indicator in relation to true malnutrition as defined by W/A and W/H are shown in Table 3. According to these findings the cut-off level of the MUAC of 13.5 cm gave high values for both sensitivity and specificity (sensitivity 71% and specificity 80 for W/A vs. sensitivity 83 and specificity 71 for W/H) for those aged 6 12 months. Cut-off levels of 14.0, 14.5 and 15.0 cm were found to be most appropriate for the ages of 13 24, and months, respectively. Table 4 shows the result of the Pearson s correlation test which was used to assess the relationship between MUAC of children and their weight, and W/A and W/H from birth to 5- years-old. All correlations had a significant association with their weight, WAZ (weight-for-age Z-score) and WHZ (weight-for-height Z-score) for all ages (P < 0.001). There were moderate positive correlations (0.50 < r < 0.75) for the majority of age groups. Strong positive correlations (r > 0.75) were found during infancy. Discussion The results of present study with regard to MUAC increments for children between 12 and 60 months are in agreement with some previous studies. 3,7,17 Mid-upper-arm circumference increases by 1 cm for boys and 1.5 cm for girls between 1 and 5 years. Compared with the NCHS median, the MUAC of children from 1- to 7-years-old were Table 3. Prevalence, sensitivity and specificity of mid-upperarm circumference (MUAC) indicators in identifying with low weight-for-height (W/H) and weight-for-age (W/A) for the 205 children from 6 to 60 months old MUAC W/A< 2SD W/H< 2SD Prevalence Sensitivity Specificity Sensitivity Specificity 6 12 months < < < < < 13.5* < months < < < < < < 14.0* < months < < < < 14.5* < < months < < < < < 15.0* < < < *Best cut-off level for children according to age group. Table 2. Prevalence of malnutrition identified by mid-upper-arm circumference (MUAC) compared with weight-for-age (W/A) and weight-for-height (W/H) for the 205 children from 6 to 60 months Indices of malnutrition (%) Age MUAC < 13.5 W/A < 2SD W/H < 2SD (months) Boys Girls Combined Boys Girls Combined Boys Girls Combined

4 68 LT Hop, R Gross, S Sastroamidjojo, T GiaY, W Schultink Table 4. Correlation (P < 0.001) between mid-upper-arm circumference (MUAC) and weight, weight-for-age Z-scores (WAZ), weight-for-height Z-Scores (WHZ) of the 205 children from 6 to 60 months old Age Correlation between MUAC and (months) Weight WAZ WHZ around 90% of the NCHS median, from 8 to 11 years about 80% and it increased after that (Table 1). It was also found that the MUAC gain of children at the age of years was relatively higher than the previous years and its standard deviation became larger. This suggests that with the beginning of puberty a new phase of growth starts (Fig. 2). From these findings it appears that the prevalence of moderate malnutrition as determined by MUAC cut-off values (13.5 cm) differs from those obtained by W/A and W/H indices. Similar results have been reported in other populations Spearman s correlation coefficient test showed that in this study increases in MUAC tends to be parallelled with weight gains of children. These result confirm that MUAC measurements allow differentiation between normal and underweight 9,10 and normal and wasted 11 children. The cut-off points of the MUAC, which have been used, are varied. Anderson 7 and Trowbridge 8 used a single cut-off point (e.g cm) for children aged 1 5 years, whereas in another study, a series of cut-off points was used to classify degrees of malnutrition. 9 In a recent study by Bern and Nathanail, 22 a higher MUAC of 14.0 cm was considered to be the most sensitive cut-off point in the detection of children with low W/H. The results of the present study indicate that a cut-off level of MUAC of 13.5 cm gave high values for both sensitivity and specificity compared with W/A and W/H indices only for children aged 6 12 months. With increasing age, sensitivity decreases, in the age group months, to only 4% in boys and 10% in girls. The data suggest that the cutoff level of MUAC 13.5 cm cannot be used in screening moderate malnutrition for all children under 5 years but should be stratified according to age group. Some tapes measure only with one marked cut-off (13.5 cm, e.g. Talc, Hertz, UK), in case they are used then only for the restricted age group (6 12 months old children). If this is not feasible for practical reasons and wider age ranges have to be assessed, then different tapes for specific age group measurements have to be used (e.g. UNICEF tape) and plotted to age. Acknowledgements. This study was financially supported by the National Institute of Nutrition, Hanoi, and the German Federal Ministry of Economic Cooperation and Development through the Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH. We are grateful to Dr Bui Thi Nhu Thuan, Dr Ha Huy Khoi and staff members of NIN for providing opportunities to undertake the study at the NIN. Our thanks also to the external consultants of GTZ and the Directorate of SEAMEO for their guidance and support to this study. We would also like to thank the participants for their kind and patient participation in this study.

5 Mid-upper-arm circumference in assessment of malnutrition 69 Mid-upper-arm circumference development and its validity in assessment of malnutrition Le Thi Hop, R Gross, S Sastroamidjojo, Tu Giay and W Schultink Asia Pacific Journal of Clinical Nutrition (1998) Volume 7, Number 1: References 1. Bray GA, Greenway FL, Molitch ME, Dahms WT, Atkinson RL, Hamilton K. Use of anthropometric measures to assess weight loss. Am J Clin Nutr 1978; 31: Harries AD, Jones LA, Heatley RV, Newcombe RG, Rhodes J. Precision of anthropometric measurements: the value of mid-arm circumference. Clin Nutr 1984; 2: Burgess HJL, Burgess AP. A modified standard for MUAC in young children. J Trop Pediatrics 1969; 15: Vijayanaghavan K, Sastry JC. The efficacy of arm circumference as a substitute for weight in assessment of protein-calorie malnutrition. An Human Biology 1976; 3: Velzeboer MI, Selwyn BJ, Sargent F, Pollitt E, Delgado H. Evaluation of arm circumference as a public health index of protein energy malnutrition in early childhood. J Trop Pediatr 1983; 29: Blankhart DM, Latham MC, Schulpen TW. Low arm circumference reporting and nutritional rehabilitation of under fives. J Trop Pediatr 1977; 23: Anderson MA. Comparison of anthropometric measurements of nutritional status in preschool children in 5 developing countries. Am J Clin Nutr 1979; 32: Trowbridge FL. Clinical and biochemical characteristics assessed with anthropometric nutritional categories. Am J Clin Nutr 1979; 321: Shakir A, Morley D. Measuring malnutrition. Lancet 1974; 1: Shakir A. Arm circumference in the surveillance of protein-calorie malnutrition in Baghdad. Am J Clin Nutr 1975; 28: McDowell I, Savage King F. Interpretation of arm circumference as an indicator of nutritional status. Arch Dis Childhood 1982; 57: Jelliffe DB, Jelliffe EFP. Arm circumference as a public health index of protein-calorie malnutrition of early childhood. J Trop Pediatrics 1969; 15: Trowbridge FL, Staehling N. Sensitivity and specificity of arm circumference indicators in identifying malnourished children. Am J Clin Nutr : WHO. Measurement of nutritional impact. WHO Geneva, WHO. Measuring change in nutritional status. WHO Geneva, Gibson RS. Anthropometric assessment of growth. In: Principles of nutritional assessment. Oxford University Press, 1990: Gayle HD, Binkin NJ, Staehling NW, Trowbridge FL. Arm circumference vs. weight-for-height in nutritional assessment: Are the findings comparable? J Trop Pediatr 1988; 34: Lindtjorn B. Measuring acute malnutrition: a need to redefine cutoff points for arm circumference? Lancet 1985; Rees DG, Henry CJK, Diskett P, Shears P. Measures of malnutrition status: survey of young children in North-East Brazil. Lancet 1987; Gurney JM, Jelliffe DB. Arm anthropometry in nutritional assessment: nomogram for rapid calculation of muscle circumference and cross-sectional muscle and fat areas. Am J Clin Nutr 1973; 26: Bern C, Nathanail L. Is mid-upper-arm circumference a useful tool for screening in emergency settings? Lancet 1995; 345: